Hart's A-frame is a five-bar exact straight-line linkage that traces a perfect rectilinear path from a coupler point without any sliding joint. Precision instrument builders rely on it where prismatic guides would introduce stiction or wear. The frame's proportions force one tracing point to move along a mathematically exact straight line as a driving point moves along an arc. The outcome is true rectilinear motion from rotary inputs alone — the same problem Harry Hart solved in 1875, still useful today for low-friction lab stages and metrology fixtures.
Hart's A-frame Interactive Calculator
Vary the critical bar length, build size, and pin clearance to see proportion error and the resulting straight-line quality.
Equation Used
The calculator follows the worked example tolerance check: compare the actual machined bar to its nominal length, express that error against the overall build size, and estimate the straight-line fuzz band from five pin clearances.
- One critical A-frame bar is compared with its nominal dimension.
- Build-size percent error follows the article's 100 mm reference comparison.
- Five pin joints are treated as a conservative linear clearance stack.
- This is a tolerance and build-quality calculator, not a full linkage synthesis solver.
The Hart's A-frame in Action
Hart's A-frame is one of two straight-line linkages Harry Hart published in 1875 — the other being his more famous six-bar Hart inversor. The A-frame uses five bars arranged so that two of them form an isoceles triangle (the 'A' shape) and the remaining three complete the kinematic chain. Pick the right proportions and one specific tracing point on the linkage moves along an exact straight line, not an approximation. No prismatic slider, no roller, no rail — just pin joints. That matters when you cannot tolerate the stiction, backlash, or wear a sliding guide introduces.
The geometry hinges on a similarity condition between the bar lengths. If the ratio between the legs of the A and the cross-link is wrong by even a fraction of a percent, the tracing point no longer lies on a straight line — it traces a shallow figure-eight or a bowed arc. We see builders get burned here all the time. A bar machined at 50.05 mm instead of 50.00 mm on a 100 mm linkage will produce visible deviation under a dial indicator. The pin joints also have to be tight; radial play in any pivot bushing of more than about 0.02 mm will smear the straight line into a fuzzy band you can measure with a height gauge.
The most common failure modes are dimensional error in the bar lengths, pin clearance accumulating across all five joints, and out-of-plane flex in the bars themselves. The linkage is planar, so any twist in a long, thin bar shows up as cross-axis wobble at the tracing point. Stiff bars, snug pivots, correct proportions — get all three right and the coupler point delivers exact straight line motion through the whole working stroke.
Key Components
- A-frame triangle (two equal bars): These two bars form the isoceles 'A' that gives the linkage its name. Their length ratio to the rest of the chain sets whether the tracing point moves on an exact line or an approximation. Match the two bars to within 0.05 mm of each other on a 100 mm build.
- Cross-link (coupler bar): Connects the apex region of the A to the input crank, completing the kinematic loop. Its length must satisfy Hart's similarity condition relative to the A-frame legs — typically a fixed fraction such as 1:2 or 2:3 depending on the variant. Get this wrong and the straight-line property is lost.
- Tracing point: The single point on the linkage whose path is the straight line. It sits at a specific intersection on one of the bars defined by the bar-length ratios. Move it 1 mm off position and the path bows.
- Pin joints (5 total): Each pivot must run with radial clearance under 0.02 mm for a precision build. Use hardened bushings or jewelled pivots in instrument-grade applications. Backlash here is the single biggest source of straight-line deviation in real builds.
- Ground pivot: The fixed pin that anchors the linkage to the frame. Its location relative to the tracing line defines the orientation of the straight-line stroke. Mount it on a stiff base — base flex translates 1:1 into tracing-point error.
- Driving crank: The input link that moves through an arc, driving the rest of the chain. The crank can rotate continuously or oscillate over a limited angle, and the tracing point will draw the same straight line for each input position within the valid range.
Industries That Rely on the Hart's A-frame
You see Hart's A-frame in places where a sliding pair simply cannot earn its keep — instrument design, metrology, classroom kinematics demonstrations, and a handful of niche industrial uses. The linkage solves the same problem the Peaucellier-Lipkin linkage solves but with five bars instead of seven. When builders pick between them they trade joint count against assembly complexity. Hart's A-frame wins on bar count; Peaucellier wins on symmetry and easier dimensioning. Both deliver exact rectilinear motion from rotary input.
- Precision metrology: Comparator stages where a probe must move along a perfectly straight reference line without the friction of a prismatic slide — used in custom-built optical flatness comparators.
- Educational kinematics: Working models in university mechanism collections, including the Reuleaux Kinematic Mechanisms collection at Cornell, which holds physical examples of Hart's straight-line linkages.
- Steam-era engineering heritage: Restoration of 19th-century beam engines and pumping stations where designers used straight-line linkages to guide piston rods, including some surviving Cornish pumping engine restorations.
- Instrument design: Low-friction tilt stages and parallel-motion drafting heads where any sliding contact would introduce unacceptable hysteresis.
- Robotics research: Specialty lab manipulators where researchers want pure rectilinear end-effector motion from revolute joints only — common in flexure-mechanism research labs at Delft and MIT.
- Watchmaking and horology: Demonstration tools and specialty escapement test rigs where a jewelled-pivot straight-line linkage outperforms any sliding guide at low loads.
The Formula Behind the Hart's A-frame
The core design equation for Hart's A-frame fixes the bar-length ratios that produce exact straight-line motion. Get this ratio right and the tracing point follows a mathematically perfect line through its working stroke. At the low end of practical bar lengths — say a 30 mm A-frame for a watchmaker's test fixture — every 0.01 mm of machining error eats roughly 0.03% of the stroke as deviation. At a nominal 100 mm build the same 0.01 mm error matters far less, around 0.01%. Push to a 500 mm linkage for a heritage engine restoration and stiffness becomes the dominant problem long before bar-length tolerance does, because thin bars flex out of plane under their own weight. The sweet spot for most lab-scale builds sits between 80 and 150 mm.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| L1 | Length of each A-frame leg (the two equal bars) | mm | in |
| L2 | Length of the cross-link (coupler bar) | mm | in |
| k | Hart similarity ratio defining the linkage proportions | dimensionless | dimensionless |
| AP | Distance from pivot A to tracing point P along the bar | mm | in |
| AB | Total bar length on which the tracing point sits | mm | in |
Worked Example: Hart's A-frame in a tabletop optical flatness comparator
A small optics workshop in Jena is building a tabletop flatness comparator that needs a probe to traverse a 40 mm straight stroke without any sliding contact. The team picks Hart's A-frame with L1 = 100 mm legs and a similarity ratio k = 0.5, giving a cross-link L2 = 50 mm. They want to verify the tracing-point position and understand how dimensional error propagates into straight-line deviation across their build tolerance range.
Given
- L1 = 100 mm
- k = 0.5 dimensionless
- Stroke target = 40 mm
- Bar tolerance (nominal) = ±0.02 mm
Solution
Step 1 — compute the cross-link length from the similarity ratio:
Step 2 — find the tracing point position along the bar:
So the tracing point sits at 0.667 × AB from pivot A, or roughly 66.7 mm along a 100 mm bar.
Step 3 — at the low end of practical build tolerance, ±0.005 mm bar accuracy (achievable on a jig-bored fixture):
That is invisible to a 1 µm dial indicator — the comparator will read straight to the limit of its instrumentation.
Step 4 — at nominal ±0.02 mm bar tolerance (typical CNC milling):
Step 5 — at the high end of acceptable tolerance, ±0.05 mm (hand-fitted bench build):
At 0.020 mm the deviation is now 50% of a typical comparator's working sensitivity. The straight line is no longer straight enough for sub-micron flatness work — the operator will read curvature in the linkage as if it were flatness error in the part. The sweet spot for this build sits around ±0.01 mm bar tolerance, achievable with a careful surface grind on the bar ends after drilling.
Result
The tracing point sits 66. 7 mm from pivot A on the 100 mm bar, with a cross-link of 50 mm completing the linkage. At nominal ±0.02 mm bar tolerance the straight line deviates by about 0.008 mm across the 40 mm stroke — fine for sub-10-µm flatness work. The low-tolerance build (±0.005 mm) drops deviation to 0.002 mm and effectively disappears into instrument noise; the high-tolerance build (±0.05 mm) bows out to 0.020 mm and starts to corrupt the measurement. If your measured deviation runs higher than 0.008 mm on a nominal build, suspect three things first: (1) unequal A-frame legs — re-measure both with a calibrated micrometer, a 0.03 mm mismatch is the single most common cause of bowed traces; (2) pin clearance over 0.02 mm at the ground pivot, which lets the whole frame shift under load; (3) out-of-plane flex in long thin bars, which shows up as the deviation changing direction when you flip the comparator on its side.
When to Use a Hart's A-frame and When Not To
Hart's A-frame is one of three exact straight-line solutions you'll consider for a precision build. The others are the Peaucellier-Lipkin linkage and Hart's own six-bar inversor. The choice comes down to bar count, joint count, dimensioning ease, and how easy the proportions are to verify on a finished build.
| Property | Hart's A-frame | Peaucellier-Lipkin | Hart inversor (6-bar) |
|---|---|---|---|
| Number of bars | 5 | 7 | 6 |
| Number of pin joints | 6 | 8 | 7 |
| Straight-line accuracy | Exact (proportions-dependent) | Exact (symmetry-dependent) | Exact (proportions-dependent) |
| Dimensioning complexity | Moderate — single ratio | Low — symmetric rhombus | High — multiple ratios |
| Sensitivity to bar tolerance | High — 0.02 mm error visible at 100 mm scale | Moderate — symmetry forgives small errors | High — multiple bars stack tolerance |
| Typical stroke length | 20-200 mm at lab scale | 10-300 mm | 20-150 mm |
| Best application fit | Compact precision instruments | Educational demos and symmetric builds | Specialty inversor research |
| Build cost (relative) | Lower — fewer parts | Higher — more parts and joints | Middle |
Frequently Asked Questions About Hart's A-frame
This is almost always coplanarity, not bar length. The linkage is planar — every joint axis must be parallel to within roughly 0.5°. If you assembled the bars with one pivot pin tipped relative to the others, the tracing point traces a 3D curve that projects onto your reference surface as a shallow arc.
Lay a precision square against each pivot pin and check perpendicularity to the base plate. A 1° tilt at one joint of a 100 mm linkage will produce roughly 0.05 mm of out-of-plane bow at the tracing point — well above your similarity-ratio error budget.
Pick Hart's A-frame when part count and packaging matter more than ease of verification. Five bars and six joints fit into a smaller footprint than Peaucellier's seven bars and eight joints.
Pick Peaucellier when you need to verify the build with a simple inspection. Peaucellier's symmetric rhombus geometry lets you check correctness with a height gauge and a pair of calipers — every rhombus side is the same length. Hart's A-frame requires you to verify a ratio between two different bar lengths, which is less forgiving when a junior tech assembles it. For lab fixtures and one-off instrument builds where the designer is also the builder, Hart's wins. For production runs assembled by multiple people, Peaucellier wins.
Useful straight-line stroke is roughly 30-50% of the longest bar — so a 100 mm linkage gives you a clean 30-50 mm working stroke. Push beyond that and the linkage approaches its singular configurations where the bars line up and the tracing point's path either reverses or jumps.
For a target stroke larger than half your longest bar, scale the whole linkage up rather than running an existing build past its sweet spot. Doubling all bar lengths doubles the stroke without changing the deviation percentage.
For most practical proportions the input link cannot rotate continuously without hitting a singular configuration where the linkage locks or snaps to its mirror pose. The straight-line property exists only over a limited input arc — typically 60-120° of crank rotation depending on the ratio you chose.
If you need continuous rotation driving a straight-line output, use a Scotch yoke or a slider-crank instead and accept the sliding pair. Hart's A-frame is fundamentally an oscillating-input mechanism. Drive it with a cam follower or a limited-arc lever, never a full-rotation crank.
That is backlash signature, and it is the most common diagnostic clue that one or more pin joints have wear or clearance issues. With 6 joints in series, even 0.01 mm of radial clearance per joint stacks up to 0.06 mm of dead-zone at the tracing point on direction reversal.
Drive the linkage with a dial indicator on the tracing point, reverse direction, and measure the dead-band. Divide by 6 — that's your average per-joint clearance. Anything over 0.02 mm per joint and you need to ream and re-bush the pivots, or move to jewelled pivots if you're working at instrument scale.
The linkage has no inherent load rating — it depends entirely on bar stiffness and pivot bearing capacity. For a typical 100 mm aluminium-bar build with 3 mm pivot pins, you can support roughly 5-10 N at the tracing point before bar deflection corrupts the straight-line accuracy.
For higher loads, scale the bar cross-section up rather than the bar length. A 6 mm × 12 mm steel bar at the same 100 mm length will carry 50 N+ at the tracing point with negligible deflection. Just remember the heavier bars also flex more under their own weight when the linkage is mounted vertically — that's a different failure mode from load-induced flex and shows up as orientation-dependent error.
References & Further Reading
- Wikipedia contributors. Straight-line mechanism. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
